As electronics evolve, there is an increased need for miniature switches that are provided on semiconductor substrates along with other semiconductor components to form various types of circuits. These miniature switches often act as relays, generally range in size from a micrometer to a millimeter, and are generally referred to as microelectromechanical system (MEMS) switches.
In some applications, MEMS switches are configured as switches and replace field effect transistors (FETs). Such MEMS switches reduce insertion losses due to added resistance, and reduce parasitic capacitance and inductance inherent in providing FET switches in a signal path. MEMS switches are currently being deployed in many radio frequency (RF) applications, such as antenna switches, load switches, transmit/receive switches, tuning switches, and the like. For instance, transmit/receive systems requiring complex RF switching capabilities may utilize a MEMS switch.
With the incorporation of WLAN (wireless local area network) circuitry into smartphones, hot switching is no longer the exception but the rule. Hot switching occurs when a switch is switched (changes state from ON to OFF, or vice versa) while some voltage exists across the switch. Specifically, many modern “smartphones” have at least two antennas: a cellular antenna and a WLAN antenna. Thus, most cycles required of a MEMS contact switch (or other switch) attached to the cellular antenna will have hot switching power due to power received from the WLAN antenna. This hot switching may be fatal to contact switches. Conversely, hot switching occurs in switches connected to the WLAN antenna due to power received from the cellular antenna.
Thus, these two antennas are presently separate but co-located in the handset such that power from one network may be coupled to the adjacent network. The attenuation between these two adjacent radios can conservatively be 5 dB. It must be assumed by component manufacturers that these two radios are in continuous operation for the lifetime of the phone, from 5 to 10 years. The peak power from a WLAN radio is known to be +28.5 dBm. The power incident to all devices directly connected to the cellular radio antenna is therefore +23.5 dBm (after subtracting the 5 dB isolation). This level of power has been shown in characterization to limit the useful lifetime of MEMS contact switches to less than 1e6 cycles. However, the required lifetime number of cycles of such a switch used in an antenna switch exceeds 1e9 cycles and is calculated to be as many as 130 B cycles.
Furthermore, MEMS contact switches are known to have a low tolerance for ESD (electro static discharge) stresses with HBM (human body model) durability not exceeding 150 V in most cases compared to a specification requirement of 250 V. No MEMS contact switch has ever been exposed to the 8 k contact discharge test common to handset ESD reliability requirements. But it is possible that a MEMS contact switch would fail this test even with the typical decoupling capacitor and shunt inductor in place.
Turning to FIGS. 1A and 1B, a MEMS device 10 having a main MEMS switch 12 is illustrated according to the prior art. The main MEMS switch 12 is formed on an appropriate substrate 14. The main MEMS switch 12 includes a cantilever 16, which is formed from a conductive material, such as gold. The cantilever 16 has a first end and a second end. The first end is coupled to the substrate 14 by an anchor 18. The first end of the cantilever 16 is also electrically coupled to a first conductive pad 20 at or near the point where the cantilever 16 is anchored to the substrate 14. Notably, the first conductive pad 20 may play a role in anchoring the first end of the cantilever 16 to the substrate 14 as depicted. The first conductive pad 20 may form a portion of or be connected to a first terminal (not shown) of the main MEMS switch 12.
The second end of the cantilever 16 forms or is provided with a cantilever contact 22, which is suspended over a terminal contact 24 formed or provided by a second conductive pad 26. The second conductive pad 26 may form a portion of or be connected to a second terminal (not shown) of the main MEMS switch 12. Thus, when the main MEMS switch 12 is actuated, the cantilever 16 moves the cantilever contact 22 into electrical contact with the terminal contact 24 of the second conductive pad 26 to electrically connect the first conductive pad 20 to the second conductive pad 26. The main MEMS switch 12 may be encapsulated by one or more encapsulating layers 30, which form a substantially hermetically sealed cavity around the cantilever 16. The cavity is generally filled with an inert gas and sealed in a near vacuum state. Once the encapsulation layers 30 are in place, an overmold 32 may be provided over the encapsulation layers 30.
To actuate the main MEMS switch 12, and in particular to cause the cantilever 16 to move the cantilever contact 22 into contact with the terminal contact 24 of the second conductive pad 26, an actuator plate 28 is formed over a portion of the substrate 14, preferably under the middle portion of the cantilever 16. To actuate the main MEMS switch 12, an electrostatic voltage is applied to the actuator plate 28. The presence of the electrostatic voltage creates an electromagnetic field that effectively moves the cantilever 16 against a restoring force toward the actuator plate 28 from an “open” position illustrated in FIG. 1A to a “closed” position illustrated in FIG. 1B. Likewise, removing the electrostatic voltage from the actuator plate 28 releases the cantilever 16 for return to the open position illustrated in FIG. 1A. As illustrated, the open position occurs when the cantilever contact 22 is out of contact with the terminal contact 24, and the closed position occurs when the cantilever contact 22 comes into contact with the terminal contact 24. Other embodiments may differ.
In light of the electromechanical structure of the main MEMS switch 12, the main MEMS switch 12 cannot provide switching action as fast as typical solid state switches, such as n-type metal-oxide-semiconductor (NMOS) FET switches. The switching time of the main MEMS switch 12 typically depends upon the electromagnetic field applied to the cantilever 16, the mass of the cantilever 16, and the restoring force of the cantilever 16. However, an FET switch may generate higher insertion loss than is generated by the main MEMS switch 12. Moreover, at high power levels in an RF circuit (not shown), parasitic capacitance at the semiconductor junctions of the FET switch may alter RF signals.
During switching events, a difference in potential between the cantilever contact 22 and the terminal contact 24 may cause an electrical arc resulting from an electrical current flowing through normally non-conductive media, such as air. Undesired or unintended electrical arcing may have detrimental effects on the cantilever contact 22 and the terminal contact 24 of the main MEMS switch 12. For instance, as the main MEMS switch 12 is being either actuated to the closed position of FIG. 1B or released to the open position of FIG. 1A, arcing from a difference in potential between the cantilever contact 22 and the terminal contact 24 may cause significant aging, unintended wear and tear, degradation, sticking, or destruction of the cantilever contact 22, the terminal contact 24, or both. Unintended power dissipation through arcing should be limited for optimum contact lifetime of the cantilever contact 22 and the terminal contact 24.
Therefore, it is evident there is a need to reduce the amount of power incident to switches connected to a cellular antenna and particularly to MEMS contact switches which are connected to a cellular antenna in order that their lifetimes can be extended to cover the lifetime cycle requirements for this application and make available their extremely low loss and high linearity.
Decreasing the amount of power incident to these switches minimizes arcing, and thus decreases switch contact aging, degradation, sticking, and destruction. In order to utilize the low loss and high linearity possible with MEMS, some means to attenuate the power must be introduced.
Furthermore, there is a need to provide some improved tolerance in these same MEMS contact switches to ESD stresses which can occur as a result of their installation at the antenna of a mobile handset.